researcharticle associationofadietaryscorewithincident...

RESEARCH ARTICLE

Association of a Dietary Score with IncidentType 2 Diabetes: The Dietary-BasedDiabetes-Risk Score (DDS)Ligia J. Dominguez1*, Maira Bes-Rastrollo2, Francisco Javier Basterra-Gortari2,3,Alfredo Gea2, Mario Barbagallo1, Miguel A. Martínez-González2

1 Geriatric Unit, Department of Internal Medicine and Geriatrics, University of Palermo, Palermo, Italy,2 Department of Preventive Medicine and Public Health, University of Navarra-IDISNA, Pamplona, Spainand CIBER Fisiopatologia de la Obesidad y Nutricion (CIBERobn), Instituto de Salud Carlos III, Madrid,Spain, 3 Department of Internal Medicine (Endocrinology), Hospital Reina Sofia, Osasunbidea-IDISNA,Tudela, Spain

* [email protected]

Abstract

Background

Strong evidence supports that dietary modifications may decrease incident type 2 diabetes

mellitus (T2DM). Numerous diabetes risk models/scores have been developed, but most do

not rely specifically on dietary variables or do not fully capture the overall dietary pattern.

We prospectively assessed the association of a dietary-based diabetes-risk score (DDS),

which integrates optimal food patterns, with the risk of developing T2DM in the SUN

(“Seguimiento Universidad de Navarra”) longitudinal study.

Methods

We assessed 17,292 participants initially free of diabetes, followed-up for a mean of 9.2

years. A validated 136-item FFQ was administered at baseline. Taking into account previ-

ous literature, the DDS positively weighted vegetables, fruit, whole cereals, nuts, coffee,

low-fat dairy, fiber, PUFA, and alcohol in moderate amounts; while it negatively weighted

red meat, processed meats and sugar-sweetened beverages. Energy-adjusted quintiles of

each item (with exception of moderate alcohol consumption that received either 0 or 5

points) were used to build the DDS (maximum: 60 points). Incident T2DM was confirmed

through additional detailed questionnaires and review of medical records of participants.

We used Cox proportional hazards models adjusted for socio-demographic and anthropo-

metric parameters, health-related habits, and clinical variables to estimate hazard ratios

(HR) of T2DM.

Results

We observed 143 T2DM confirmed cases during follow-up. Better baseline conformity with

the DDS was associated with lower incidence of T2DM (multivariable-adjusted HR for inter-

mediate (25–39 points) vs. low (11–24) category 0.43 [95% confidence interval (CI) 0.21,

PLOS ONE | DOI:10.1371/journal.pone.0141760 November 6, 2015 1 / 15

OPEN ACCESS

Citation: Dominguez LJ, Bes-Rastrollo M, Basterra-Gortari FJ, Gea A, Barbagallo M, Martínez-GonzálezMA (2015) Association of a Dietary Score withIncident Type 2 Diabetes: The Dietary-BasedDiabetes-Risk Score (DDS). PLoS ONE 10(11):e0141760. doi:10.1371/journal.pone.0141760

Editor: Manlio Vinciguerra, University CollegeLondon, UNITED KINGDOM

Received: June 25, 2015

Accepted: October 13, 2015

Published: November 6, 2015

Copyright: © 2015 Dominguez et al. This is an openaccess article distributed under the terms of theCreative Commons Attribution License, which permitsunrestricted use, distribution, and reproduction in anymedium, provided the original author and source arecredited.

Data Availability Statement: Data contain potentiallyidentifying information and are available upon requestto the corresponding author.

Funding: The Seguimiento Universidad de Navarra(SUN) study has received funding from the SpanishMinistry of Health and European RegionalDevelopment Fund (FEDER) (grants PI10/02993,PI10/02658, PI13/00615, PI14/01668, PI14/01798,PI14/1764, RD06/0045, G03/140), and the NavarraRegional Government (45/2011, 122/2014). AG issupported by a FPU fellowship from the SpanishGovernment.

http://crossmark.crossref.org/dialog/?doi=10.1371/journal.pone.0141760&domain=pdf

http://creativecommons.org/licenses/by/4.0/

0.89]; and for high (40–60) vs. low category 0.32 [95% CI: 0.14, 0.69]; p for linear trend:

0.019).

Conclusions

The DDS, a simple score exclusively based on dietary components, showed a strong

inverse association with incident T2DM. This score may be applicable in clinical practice to

improve dietary habits of subjects at high risk of T2DM and also as an educational tool for

laypeople to help them in self-assessing their future risk for developing diabetes.

IntroductionType 2 diabetes mellitus (T2DM) is a pandemia of this century. This common chronic diseasehas increased massively in recent years in parallel with the obesity epidemic, affecting 382 mil-lion people worldwide in 2013 [1]. If the same trend continues, the estimates foresee that by2035 there will be 592 million persons with T2DM, particularly among young adults and citi-zens from low- and middle-income countries [1–3]. These numbers are worrisome becauseT2DM is associated with substantial increased risk of cardio- and cerebrovascular events andmortality [4], as well as with severe disability due to blindness [5], chronic renal failure [6], andlower limb amputations [7]. The resulting mortality and disability entail overwhelminghuman, financial, and social burden [1–3]. The cost of preventing and treating T2DMmay behigh, but the cost of neglecting it will be vastly higher.

There is strong evidence demonstrating that dietary modifications may decrease incidentT2DM by 33% [8], 50% [9], and 58% [10] in people at high-risk from China, Finland, and theUSA, respectively. Moreover, the effect of the implemented dietary interventions in these threestudies persisted in the long term [11–13]. Two recent systematic reviews evaluated availableevidence for effectiveness including 53 studies [14] and cost-effectiveness including 28 studies[15] of combined diet and physical activity promotion programs concluding that these pro-grams are effective in reducing new-onset T2DM, increasing reversion to normoglycemia, andimproving diabetes and cardiometabolic risk factors in persons at risk. Programs that achieveda mean weight loss at one year of only 2.5% resulted in 60% reduction in diabetes developmentat 6 years [15].

Hence, it is essential to detect persons at risk for diabetes to implement intensive preventiveinterventions the earliest possible. Numerous diabetes risk models and scores have been devel-oped but most are rarely used. A systematic review evaluated in detail 94 such models, testing6.88 million participants followed for up to 28 years. Heterogeneity of the studies included inthe review precluded meta-analysis. These predictive scores do not rely specifically on dietaryvariables or only include few nutritional items. Some scores may involve biochemical analyticaltests not regularly available; other scores searching simplicity and practicality lack of complete-ness. Most scores/models are not focused on diet, include only few and general food compo-nents and do not fully capture the overall dietary pattern [16]. This is in contrast to thefundamental role of dietary habits as key determinants of obesity and T2DM [17].

Prospective cohort studies have provided evidence on the contribution of several specificdietary factors in the development of T2DM [18–20]. However, the total effect of all these die-tary factors together has not been jointly evaluated to build an a priori dietary-based diabetesrisk score. Therefore, we conducted the present analyses aiming to evaluate a dietary-based

Dietary Score and Incident Diabetes


Competing Interests: The authors have declaredthat no competing interests exist.

diabetes-risk score (DDS), and its association with incident T2DM, using data from the Medi-terranean cohort of the SUN −Seguimiento Universidad de Navarra−project.

Methods

Study design and populationThe SUN project is a prospective, permanently open, dynamic cohort of university graduatesstarted in 1999 with biennial collection of updated information. The design and methods of theSUN study have been previously described in detail and can be found elsewhere [21,22]. TheInstitutional Review Board of the University of Navarra approved the study protocol. The ini-tial response to a mailed questionnaire was considered as informed consent to participate.

For the present analyses, we examined the last available database as of December 2014, cor-responding to 22,175 participants. We included participants who had spent enough time in thestudy (>2 years and additional 9 months) as to be able to complete and return at least the2-year follow-up questionnaire; otherwise, they were excluded (n = 3108). Participants wereexcluded from the analyses if they reported total energy intake out of pre-defined limits [23](n = 2089), or had a previous diagnosis of diabetes (n = 404). Some of them had more than oneexclusion criteria. The final analytic population included 17,292 participants. Those with miss-ing values in smoking (n = 575) were treated as another category (current/former smoker/never smokers/missing). Overall retention was 93.1% (93.1% of participants recruited at least2 years and 9 months ago returned�1 of the follow-up questionnaires).

Dietary assessmentDietary habits were assessed at baseline by a semi-quantitative 136-item FFQ previouslydescribed in detail [24]. The validity [24,25] and reproducibility [26] of this questionnaire havebeen repeatedly reported. Nutrient scores were computed as previously described in detail [24–26] using the latest available Spanish food composition tables [27,28]. Table 1 shows the foodsand nutrients included in the diabetes score.

To build the dietary-based diabetes-risk score (DDS), we considered the consumption (g/d)of nine nutritional exposures which have shown associations with a decreased incidence ofT2DM (vegetables, fruit, fiber, whole cereals, nuts, coffee, PUFA, low-fat dairy, alcohol in mod-erate amounts), and three food groups which have shown associations with an increased inci-dence of T2DM (red meat, processed meat, and sugar-sweetened beverages [SSB]) [18–20]. Weadjusted the consumption of each nutritional variable for total energy intake by using the resid-ual method separately for men and women [29]. The energy-adjusted estimates (residuals)were ranked according to their sex-specific quintile values (assigning a value of 1 for the firstquintile, 2 for the second quintile, and successively until the value of 5 was assigned to the fifthquintile). The quintile values for the food groups with increased risk of incident T2DM werereversed (assigning a value of 5 for the first quintile, 4 for the second quintile, and successivelyuntil the value of 1 was assigned to the fifth quintile). For alcohol, 5 points were assigned formoderate consumption (10–50 g/day for men and 5–25 g/day for women); otherwise, the par-ticipant received zero points. To obtain the DDS, quintile values of nutritional variables withexpected protection and reverse quintile values of food groups with expected increased riskwere summed; thus, the final scores could range from 11 (lowest adherence) to 60 (highestadherence) points. We classified adherence to the DDS in 3 categories: low (11–24), intermedi-ate (25–39), and high (40–60). We chose to use these specific round cut-offs points instead ofquantiles because the groups thus built are more meaningful per se and could be more easilyused for future comparisons with similar studies. This is in line with current recommendationsgiven in epidemiology about procedures to categorize continuous variables [30]. Additionally,



we also show the results of alternative analyses using absolute consumption cutoffs of themodel (see below).

As sensitivity analyses, we built a similar model but considering the score as a continuousvariable (for one and five additional points). We also repeated the analyses for men andwomen, separately; older (>50 years) and younger (<50 years) persons; higher (�30 kg/m2)and lower (<30 kg/m2) BMI; changing the energy limits, including only participants withenergy between percentile 1 and 99; and, excluding cases with an early diagnosis of T2DM dur-ing follow-up. We also built a similar score with the same exposures but using absolute cut-offspoints for each one of the 12 food groups with goals expressed as servings/day or servings/week (i.e., normative or absolute cutoffs) [31,32] instead of using energy-adjusted categories ofconsumption. However, the same food groups as in the category-based assessment were con-sidered and the score assigned 1 point for each of the 12 goals that was met (consumption ofeach item at or above the limit for protective foods and below the limit for deleterious foods).The cut-off points were as follows: vegetables (�2/d), fruit (�2/d), whole bread (>0/d), fiber(�25 g/d), coffee (�3/d), nuts (�3/d), low-fat dairy products (�1/d), PUFA (�5 g/d), alcohol(>10 g<50 g/d for men;>5 g<25 g/d for women), meat (�1/d), processed meat (�1/d), SSB(�1/d). In these analyses, we did not compute residuals from regressions on energy intake;instead, we adjusted for total energy intake by introducing it as a covariate in the standard mul-tivariable models.

Ascertainment of DiabetesDetailed information on ascertainment of T2DM in the SUN cohort has been reported before[33]. In brief, we considered diabetes at baseline if participants reported a medical diagnosis orwere receiving oral antidiabetic agents or insulin. We considered probable incident cases forparticipants who reported a T2DM diagnosis made by a doctor during follow-up. We sent

Table 1. Scoring criteria for the Diabetes Dietary Score in the SUN cohort, 1999–2014.

Component Included foods

Protection• Vegetables Carrot, pumpkin, Swiss chard, cabbage, cauliflower, broccoli, lettuce,

chicory, escarole, tomatoes, green beans, eggplant, zucchini, cucumber,peppers, asparagus, spinach, other fresh vegetables

• Fruit Citrus, banana, apple, pear, strawberry, peach, cherry, fig, melon,watermelon, grapes, kiwi, canned fruit

• Total dietary fiber g/day

• Whole cereals whole-grain bread

• Nuts Almonds, peanuts, hazelnuts, pistachios, pine nuts, walnuts

• Coffee Cups (50 ml) of coffee consumed

• PUFA g/day

• Low-fat dairy Skim or low-fat milk

• Alcohol(moderateconsumption)

10 g/day for men and 5 g/day for women

Increased risk• Red meat Beef, pork, lamb, liver

• Processed meat Cooked ham, Parma ham, mortadella, salami, foie gras, spicy porksausage, bacon, cured meats, hamburger, hot dog

• SSB All sugar sweetened beverages

PUFA: Polyunsaturated Fatty Acids; SSB: sugar sweetened beverages

doi:10.1371/journal.pone.0141760.t001



additional specific questionnaires to these participants to confirm their diagnosis, and to spec-ify further details (i.e., type, date of diagnosis, gestational diabetes, highest fasting glucosevalue, eventual OGTT, HbA1c, current use of oral antidiabetic agents or insulin, and occur-rence of complications). Probable cases were requested to send us their medical reports detail-ing the diagnosis. A panel of physicians, blinded to dietary habits information, classified thesemedical records and adjudicated the cases as confirmed incident T2DM or not. As previouslyreported [33], the diagnosis criteria for confirmed T2DM cases were those of the AmericanDiabetes Association.

Other covariatesCovariates assessed at baseline included socio-demographic parameters (age, marital status,years of university education), anthropometric measurements (weight, BMI), health-relatedhabits (smoking status, physical activity, sedentary lifestyle, hours sitting down, hours of televi-sion watching), and clinical variables (medications, personal history of hypertension, familyhistory of diabetes). Self-reported weight and BMI have been previously validated in a sub-sample of this cohort [34]. Physical activity was assessed using a previously validated question-naire with a Spearman correlation coefficient of 0.51 (p<0.001) with objective measurements[35]. Physical activity was expressed in metabolic equivalent tasks (METs-h/week) as calculatedfrom the time spent at each activity in hours/week multiplied by its typical energy expenditure[36]. Adherence to the Mediterranean food pattern was appraised using the score proposed byTrichopoulou [37].

Statistical analysisFor building the DDS, we used only the information from the baseline FFQ. Means with SDsfor continuous baseline characteristics and proportions for categorical characteristics were cal-culated by categories of DDS. The time to the event was defined as the number of days fromrecruitment to the last questionnaire, death or a confirmed diabetes diagnosis as determined bythe adjudicator, whichever came first. Cox proportional hazards analyses were fitted to assessthe association of the DDS with incident T2DM. After a crude analysis, we fitted a modeladjusted for sex and age (as the underlying time variable). In a subsequent model we addition-ally adjusted for major risk factors of T2DM (total energy intake, adoption of special diets,snacking between meals, baseline BMI, physical activity, hours of television watching, hourssitting down, smoking, marital status, personal history of hypertension, and family history ofdiabetes). Robust standard errors were used. All models were stratified by age groups and yearof recruitment. The p for trend was calculated taking the median for each category and intro-ducing this new variable as a continuous variable in the models. We evaluated the interactionbetween the dietary score and BMI through the likelihood ratio test for the fully-adjustedmodel with and without the product-term. Nested regression models after a stepwise forwardselection algorithm were used to evaluate the contribution of each item to the final score. Asthe BMI is also related to dietary habits, we also evaluated the association between baselineBMI and the risk of T2DM in an ancillary analysis, using similar methods.

The analyses were performed with Stata software package version 12 (Stata Corp). All testswere two sided and statistical significance was set at P<0.05.

ResultsTable 2 shows baseline characteristics of the studied population of participants in the SUNproject, according to categories of DDS (low-to-high). Older participants, married participants,those with higher university years of education and more physically active participants were



Table 2. Baseline categories of participants according to categories of the Diabetes Dietary Score (DDS) in the SUN (“Seguimiento Universidad deNavarra”) cohort, 1999–2014.

Diabetes Dietary score (DDS)

Low (11–24) Intermediate (25–39) High (40–60) p1

N 1180 12076 3893

Sex, male (%) 40.3 39.6 43.1 0.001

Age (y) 32.1 ± 9.12 37.8 ± 11.3 43.2 ± 12.4 <0.001

Marital status (married) (%) 38.7 50.9 58.6 <0.001

Years of university education 4.8 ± 1.3 5.1 ± 1.5 5.2 ± 1.6 <0.001

Family history of diabetes (%) 11.0 15.1 19.1 <0.001

Hypertension (%) 3.3 6.8 10.7 <0.001

High blood cholesterol (%) 8.2 16.1 24.1 <0.001

BMI (kg/m2) 22.7 ± 3.3 23.6 ± 3.5 23.9 ± 3.5 <0.001

Smoking (%)

Current 25.3 22.5 20.0 <0.001

Former smoker 18.8 27.5 40.1

Alcohol intake (g/d) 5.0 ± 13.3 6.2 ± 10.1 8.7 ± 9.4 <0.001

Physical activity (MET-h/wk) 18.6 ± 22.3 20.5 ± 21.6 25.8 ± 24.8 <0.001

Hours sitting down/d 5.9 ± 2.0 5.7 ± 1.9 5.5 ± 2.0 <0.001

TV watching (h/d) 4.5 ± 2.8 4.7 ± 2.7 4.7 ± 2.7 0.26

Vegetables (g/d) 320 ± 208 482 ± 297 720 ± 38 <0.001

Fruit (g/d) 183 ± 141 310 ± 271 497 ± 342 <0.001

Legumes (g/d) 22 ± 16 22 ± 18 24 ± 18 <0.001

Cereals (g/d) 122 ± 83 100 ± 73 100± 67 <0.001

Whole bread (g/d) 1.3 ± 9.1 9.4 ± 27 28 ± 42 <0.001

Potatoes (g/d) 27 ± 29 27 ± 29 29 ± 31 <0.01

Nuts (g/d) 3.4 ± 4.2 5.8 ± 8.9 13 ± 18 <0.001

Olive oil (g/d) 18 ± 15 18 ± 15 20 ± 15 <0.001

Meats/meat products (g/d) 250 ± 85 180 ± 73 132 ± 67 <0.001

Animal fats for cooking or as a spread (g/d) 1.7 ± 3.4 1.1 ± 2.7 0.7 ± 2.0 <0.001

Eggs (g/d) 28 ± 19 24 ± 16 20 ± 13 <0.001

Fish and other seafood (g/d) 84 ± 53 94 ± 58 114 ± 68 <0.001

Whole dairy products (g/d) 360 ± 251 204 ± 195 116 ± 138 <0.001

Low-fat dairy products (g/d) 97 ± 179 212 ± 240 308 ± 260 <0.001

Coffee (cups/d) 0.76 ± 1.1 1.16 ± 1.2 1.53 ± 1.3 <0.001

Following a special diet 2.7 6.9 14.0 <0.001

Between-meal snacking 44.0 33.6 26.2 <0.001

Dietary intakes

Total energy (kcal/d) 2703 ± 544 2328 ± 616 2291 ± 610 <0.001

Carbohydrate (% of energy) 41 ± 7.2 43 ± 7.3 45 ± 7.6 <0.001

Protein (% of energy) 17.8 ± 3.2 18.2 ± 3.3 18.1 ± 3.3 0.002

Total fat (% of energy) 40 ± 6.0 37 ± 6.3 34 ± 6.6 <0.001

MUFAs (% of energy) 17 ± 3.3 16 ± 3.7 15 ± 3.8 <0.001

SFAs (% of energy) 15 ± 3.1 13 ± 3.0 10 ± 2.8 <0.001

PUFAs (% of energy) 5.1 ± 1.5 5.2 ± 1.5 5.1 ± 1.4 0.007

Vitamin C (mg/d) 187 ± 86 258 ± 138 374 ± 178 <0.001

Vitamin D (mcg/d) 3.5 ± 2.2 3.6 ± 2.4 4.2 ± 2.8 <0.001

Iron from heme sources (mg/d) 16 ± 4 16 ± 5 19 ± 6 <0.001

Folate (mcg/d) 307 ± 108 383 ± 158 521 ± 194 <0.001

(Continued)



more likely to belong to the highest category of the DDS; whereas current smokers, and thosewith a higher total energy intake were more likely to belong to the lowest category of the DDS(Table 2). As expected, the consumption of the nine favorable nutritional factors (with theexception of PUFA intake) increased monotonically across increasing categories of the DDS,whereas the consumption of the three detrimental food groups monotonically decreased(P<0.001 for all) (Table 2). The most striking differences were observed for the consumptionof whole bread, nuts (including all tree nuts and peanuts), low-fat dairy, fruits, and vegetables.The intakes of vitamin C, heme iron from heme sources, folate, and fiber were greater in thehigh score category group. Conversely, the intakes of total energy and total fat were lower inthe high score category group. Although significant, only minimal differences were observedfor carbohydrate and vitamin D intakes across DDS categories (Table 2). Analyses of the con-tribution of the different components of the DDS with nested regressions after a stepwise selec-tion algorithm showed that the largest variability was explained by vegetables (18%) and low-fat dairy consumption (13%), with no single component apparently driving the score.

During 159,567 person-years follow-up (mean follow-up: 9.2 years; range: 1.6–15.3 years),we confirmed 143 first diagnoses of T2DM among 17,292 participants of the SUN project. Asignificantly inverse linear trend in the Cox model was apparent for the association betweenDDS and the risk of T2DM after adjustment for sex and age (P = 0.004) (Table 3). The fully-adjusted HRs (95% CIs) for categories of intermediate and high adherence compared with thelow adherence category (reference) were 0.43 (0.21, 0.89), and 0.32 (0.14, 0.69), respectively,with a significant inverse linear trend (P = 0.019) (Table 3). The DDS and the Mediterranean

Table 2. (Continued)

Diabetes Dietary score (DDS)

Low (11–24) Intermediate (25–39) High (40–60) p1

Dietary fiber (g/d) 20 ± 7 26 ± 11 36 ± 14 <0.001

MET: metabolic equivalent task; MUFA: monounsaturated fatty acid; SFA: saturated fatty acid; PUFA: polyunsaturated fatty acid1 Comparisons of characteristics across categories of the diabetes dietary score were performed by using 1-factor ANOVA for quantitative variables or chi-

square tests for categorical variables.2 Mean ± SD (all such values)


Table 3. HRs (95% CIs) of incident diabetes according to baseline categories of the Diabetes Dietary Score in the SUN cohort, 1999–20141.

Diabetes Dietary Score

Low (11–24) Intermediate (25–39) High (40–60) p

N 1180 12076 3893

n of incident diabetes 10 99 34

Persons-year of follow-up 11793 113284 34490

Age-adjusted diabetes incidence (x 10−3) 0.85 (0.41, 1.56) 0.45 (0.23, 0.86) 0.32 (0.16, 0.67)

Crude HR* 1 (ref) 0.47 (0.24, 0.93) 0.32 (0.15, 0.66) 0.006

Age-, year of recruitment and sex-adjusted HR 1 (ref) 0.46 (0.23, 0.91) 0.30 (0.15, 0.64) 0.004

Multivariate-adjusted HR1 1 (ref) 0.43 (0.21, 0.89) 0.32 (0.14, 0.69) 0.019

* Age as the underlying time variable.1 Adjusted for sex, total energy intake, following a special diet, snacking between meals, BMI, physical activity, hours of television watching, hours sitting

down, smoking, marital status, personal history of hypertension, and family history of diabetes (parents and/or siblings).




diet score [37] had a Spearman correlation coefficient of 0.617 (p<0.001), suggesting a moder-ate degree of overlapping between both a priori-built dietary scores.

In the ancillary analysis of the association between BMI and T2DM, we found that aBMI>35 comported a 64-fold increase in incident diabetes compared to a BMI<22 (Fig 1);this strong effect persisted when considering separately men (25 fold) and women (56 fold),with higher increases for women, and when considering participants younger (69.5 fold) andolder than 50 years (59 fold). There was a continuous rise in the risk of diabetes as the BMIincreased. Interestingly, the increased risk (4 fold) was already evident when comparing per-sons with BMI from 22.1 to 24.9 versus those with BMI<22 (Fig 1).

Several sensitivity analyses were carried out in order to appraise the strength of our findings(Fig 2). Assessing DDS as a continuous variable, for each five additional points, the risk of dia-betes decreased by 15%. When we assessed men and women separately, for both groups thedietary score was inversely associated with T2DM risk in multiple-adjusted models; the com-parison of high vs. low category was significant for men, but not for women. Considering par-ticipants older and younger than 50 years separately, for both groups the fully-adjusted DDSshowed an inverse association; the comparison of the high vs. low category was significant forparticipants older than 50 years, but it was not significant for participants younger than 50years. Examining separately participants with BMI higher and lower than 30, for those withBMI�30 comparison of the high vs. low category was significant, while it was also inverselyassociated with T2DM for those with BMI<30 without reaching statistical significance. No sig-nificant interaction was observed between the score and BMI, when we dichotomized BMI by

Fig 1. Risk of developing T2DM and BMI.Multivariable-adjusted hazard ratios for the risk of developing T2DM according to baseline body mass index. TheSUN cohort 1999–2014.

doi:10.1371/journal.pone.0141760.g001



30 kg/m2, or when we considered BMI as a continuous variable in the interaction product-term. After excluding early cases of T2DM (those diagnosed during the first 2-year follow-upperiod), the score was still inversely associated with T2DM, but without reaching statisticalsignificance.

Finally, we constructed a similar diabetes dietary score using servings/day or servings/week(i.e., normative or absolute cut-off points [31,32] instead of using energy-adjusted categories ofconsumption). Because this score assigned one point to each of the 12 goals accomplished, itspossible range was 0–12 points. The fully-adjusted HRs in this sensitivity analysis is shown inFig 3.

DiscussionWe assessed an a priori dietary-based diabetes score (the DDS) to appraise the association ofthe total dietary pattern with type 2 diabetes. This score, in contrast with previous models, wasbased on several specific dietary components with available previous evidence of their associa-tion with increased or decreased risk of T2DM. This is important because of the key role of die-tary habits as determinants of obesity and T2DM. It may represent a useful tool, not only foridentification of high-risk individuals according to their dietary pattern, but also to educateconsumers on healthy dietary and lifestyle choices while assessing their risk of diabetes.

To construct our DDS, we used dietary factors proved to contribute to T2DM risk [18–20].Therefore, there was a rationale to incorporate these elements to build an a priori dietarymodel. These factors represent established risk parameters, but some of them have not beencombined in a single score nor even included in previous diabetes risk assessment models [16].

Fig 2. Risk of developing T2DM and DDS. Sensitivity analyses of the association between the dietary score and the risk of incident diabetes. Multivariableadjusted model. The SUN Project 1999–2014.




While only modifiable risk factors can be addressed by interventions, non-modifiable risk fac-tors, like age and family history of T2DM, contribute significantly to determine a person’s riskand were all included in our analyses as potential confounders.

We found a significant, but moderate correlation between the DDS and the Mediterraneandiet score (MDS). Although the MDS was not built to combine nutritional factors reported tobe associated with T2DM, it has been consistently associated with lower incidence of T2DM[31,33,38]. Some of the components are shared by the DDS and the MDS, but, there are somedifferences between both scores. For example, we did not include legumes in the DDS becausethere is still no report showing relevant association of legumes intake with incident diabetes.Instead of including cereals we included total fiber and whole grains because both have beenassociated with a reduced diabetes risk [18,39]. Dairy products are considered detrimental inthe MDS, while there is evidence supporting an inverse association of some dairy products,especially low-fat dairy, with diabetes risk [18,19], which we included in the DDS. We alsoincluded PUFA and coffee intake because both were reported to be associated with lower

Fig 3. Hazard ratios (HR) and 95% confidence interval (CI) for incident T2DM according to a DDS. For this analysis we used servings/day or servings/week (i.e. normative or absolute cut-off points [31,32]), assigning one point to each of the 12 goals accomplished. Results represent a fully adjusted model,adjusted for age (as the underlying time variable), sex, total energy intake, following a special diet, snacking between meals, BMI, physical activity, hours oftelevision watching, hours sitting down, smoking, marital status, personal history of hypertension, and family history of diabetes [parents and/or siblings]).




diabetes risk [18,20]; SSB are not considered in the MDS, but there is evidence on their associa-tion with an increased risk of diabetes [18].

A number of diabetes risk scores are now available providing a fairly good but not perfectestimate of the probability to develop diabetes in the years ahead. In a review of 94 T2DM riskassessment models, only seven were suitable for clinical practice [16]. Most models claimedwide applicability, which is difficult to reconcile with the inherent selection bias of the sample/cohort characteristics used to develop many scores. It appears that there is a recent shift in pri-orities from the exclusive chase of statistical brilliance to the practical application and out-comes of using diabetes risk scores in real-world prevention programs [16,40]. We aimed toassess whether a score exclusively based on the potential role of nutritional elements may showa strong association with incident T2DM. Probably, there is no single ideal risk score univer-sally applicable, as the value depends not only on its statistical properties but also on its contextof use, which may define which data are available to consider. Perhaps development ofnational/regional models and risk categories, according to the local/regional characteristics andtraditions is advisable. Extrapolation of scores in different contexts/populations showed vari-ability in estimates of risk as high as twenty-fold [41,42]. The model we propose is inexpensive,can be self-administered, and it is directed to educate laypeople and to encourage persons atrisk to adopt/improve their healthy dietary choices. These healthy choices, proved to preventT2DM, may also influence the incidence of other non-communicable diseases [43], therefore, awidespread dissemination of these type of self-assessment is warranted.

Several RCTs [8–14,44] have shown that weight reduction should be the primary goal ofdiet and lifestyle interventions addressed to prevent T2DM. In the Nurses’Health Study,T2DM risk over a 14-year follow-up was 49-fold higher among women with baseline BMI>35vs. those with BMI<22 [45]. Participants of the US male health professionals cohort withBMI�35 had RR of incident diabetes 42-fold higher than those with BMI<23, after adjustmentfor confounders [46]. We found an even higher relative risk of T2DM according to BMI in ouryounger SUN cohort, with a 64-fold increase in incident diabetes for BMI>35 vs BMI<22.However, aging is a substantial risk marker and the absolute risk will be always higher with age[47]. Importantly, the risk of diabetes was increased by*4-fold for persons with BMI from22.1 to 24.9 vs. BMI<22, emphasizing the key role of achieving and maintaining an ideal BMIas early as possible and highlighting the perils of a slightly increased BMI within the normalrange. We included this assessment together with the DDS in order to provide a perspective onthe relative roles of dietary composition and BMI on the risk of developing T2DM. Althoughbeing too thin or losing weight rapidly is associated with higher mortality risk among older per-sons [48], a recent study showed no evidence of protection for overweight/obesity on mortalityin older persons with T2DM who never smoked [49].

The strengths of our study are: a) large sample size; b) high retention rate; c) prospectivedesign; d) lengthy follow-up; e) ability to control for a wide array of confounding factors,including potential lifestyle and demographic confounders; f) inclusion of several sensitivityanalyses where the results still pointed to a negative association in their point estimates (thoughconfidence intervals were wide because of the reduction in sample size caused by splitting thesample).

Potential limitations include: a) self-reported information, however, parameters such asself-reported weight and BMI and the FFQ have been previously validated in sub-samples ofthis cohort [34]; b) the cohort is composed of middle-aged, highly educated persons, with lowprevalence of overweight/obesity and high level of physical activity. This explains the relativelylow number of observed cases and the consequent width of some confidence intervals. How-ever, we found strongly significant inverse associations for DDS and strongly significant directassociations for BMI; c) though SUN cohort is composed of highly educated participants,



regardless of this fact, the generalizability of our results must be based on common biologicalmechanisms instead of on statistical representativeness; we used restriction to reduce potentialconfounding by disease, education, socioeconomic status, and presumed access to health care;however, future studies are needed in order to test the applicability of our results to other popu-lations; d) finally, concerns may arise from the use of FFQ, which may be subject to informa-tion bias. However, the FFQ used has been repeatedly validated [24–26]; furthermore, it isdifficult to find a better alternative method to characterize food habits of large samples of per-sons, followed over long periods of time, in order to assess associations with incident clinicalend-points [23].

In conclusion, a score exclusively based on dietary components showed a strong inverseassociation with incident T2DM. This score may be applicable in clinical practice because it isbased on variables that can be gathered in primary care, and this score may be even gatheredusing self-administered tools. Furthermore, it may well be an educational tool for laypeople toself-assess their risk of diabetes. Future studies are warranted in order to test whether the appli-cation of this model may be able to help modify dietary choices, incident T2DM and relatedmorbidity.

Supporting InformationS1 File. STROBE checklist for cohort studies.(DOC)

AcknowledgmentsThe authors thank the participants of the SUN Project for their enthusiastic collaboration.

Author ContributionsConceived and designed the experiments: LJD MAM-G. Performed the experiments: MAM-GFJB-G MB-R AG. Analyzed the data: MAM-G LJD. Wrote the paper: LJD MAM-GMB-R MBFJB-G AG.

References1. International Diabetes Federation. IDF Diabetes Atlas (6th edition) (2013) Available at: http://www.idf.

org/diabetesatlas. Accessed 5 March 2015.

2. Zimmet PZ, Magliano DJ, HermanWH, Shaw JE. Diabetes: a 21st century challenge. Lancet DiabetesEndocrinol. 2015; 2:56−64.

3. Seaquist ER. Addressing the burden of diabetes. JAMA. 2014; 311:2267−2268. doi: 10.1001/jama.2014.6451 PMID: 24915253

4. Joseph JJ, Golden SH. Type 2 diabetes and cardiovascular disease: what next? Curr Opin EndocrinolDiabetes Obes. 2014; 21:109−120. doi: 10.1097/MED.0000000000000044 PMID: 24569552

5. Ciulla TA, Amador AG, Zinman B. Diabetic retinopathy and diabetic macular edema. Diabetes Care.2003; 26:2653−2664. PMID: 12941734

6. Jha V, Garcia-Garcia G, Iseki K, Li Z, Naicker S, Plattner B, et al. Chronic kidney disease: global dimen-sion and perspectives. Lancet. 2013; 382:260−272. doi: 10.1016/S0140-6736(13)60687-X PMID:23727169

7. DiPreta JA. Outpatient assessment and management of the diabetic foot. Med Clin North Am. 2014;98:353−373. doi: 10.1016/j.mcna.2013.10.010 PMID: 24559880

8. Pan XR, Li GW, Hu YH, Wang JX, YangWY, An ZX, et al. Effects of diet and exercise in preventingNIDDM in people with impaired glucose tolerance. The Da Qing IGT and Diabetes Study. DiabetesCare. 1997; 20:537−544. PMID: 9096977



http://www.plosone.org/article/fetchSingleRepresentation.action?uri=info:doi/10.1371/journal.pone.0141760.s001

http://www.idf.org/diabetesatlas

http://www.idf.org/diabetesatlas

http://dx.doi.org/10.1001/jama.2014.6451


http://www.ncbi.nlm.nih.gov/pubmed/24915253

http://dx.doi.org/10.1097/MED.0000000000000044



http://dx.doi.org/10.1016/S0140-6736(13)60687-X


http://dx.doi.org/10.1016/j.mcna.2013.10.010



9. Tuomilehto J, Lindstrom J, Eriksson JG, Valle TT, Hämäläinen H, Ilanne-Parikka P, et al. Finnish Diabe-tes Prevention Study Group. Prevention of type 2 diabetes mellitus by changes in lifestyle among sub-jects with impaired glucose tolerance. N Engl J Med. 2001; 344:1343−1350. PMID: 11333990

10. Knowler WC, Barrett-Connor E, Fowler SE, Hamman RF, Lachin JM, Walker EA, et al. Diabetes Pre-vention Program Research Group. Reduction in the incidence of type 2 diabetes with lifestyle interven-tion or metformin. N Engl J Med. 2002; 346:393−403. PMID: 11832527

11. Li G, Zhang P, Wang J, An Y, Gong Q, Gregg EW, et al. Cardiovascular mortality, all-cause mortality,and diabetes incidence after lifestyle intervention for people with impaired glucose tolerance in the DaQing Diabetes Prevention Study: a 23-year follow-up study. Lancet Diabetes Endocrinol. 2014; 2:474−480. doi: 10.1016/S2213-8587(14)70057-9 PMID: 24731674

12. Lindstrom J, Ilanne-Parikka P, Peltonen M, Aunola S, Eriksson JG, Hemio K, et al. Finnish DiabetesPrevention Study Group. Sustained reduction in the incidence of type 2 diabetes by lifestyle interven-tion: follow-up of the Finnish Diabetes Prevention Study. Lancet. 2006; 368:1673−1679. PMID:17098085

13. Knowler WC, Fowler SE, Hamman RF, Christophi CA, Hoffman HJ, Brenneman AT, et al. Diabetes Pre-vention Program Research Group. 10-year follow-up of diabetes incidence and weight loss in the Dia-betes Prevention Program Outcomes Study. Lancet. 2009; 374:1677−1686. doi: 10.1016/S0140-6736(09)61457-4 PMID: 19878986

14. Balk EM, Early A, Raman G, Avendano EA, Pittas AG, Remington PL. Combined diet and physicalactivity promotion programs to prevent type 2 diabetes among persons at increased Risk: a systematicreview for the Community Preventive Services Task Force. Ann Intern Med. 2015; 163 [Epub ahead ofprint 14 July 2015].

15. Li R, Qu S, Zhang P, Chattopadhyay S, Gregg EW, Albright A, et al. Economic evaluation of combineddiet and physical activity promotion programs to prevent type 2 diabetes among persons at increasedrisk: a systematic review for the Community Preventive Services Task Force. Ann Intern Med. 2015;163 [Epub ahead of print 14 July 2015].

16. Noble D, Mathur R, Dent T, Meads C, Greenhalgh T. Risk models and scores for type 2 diabetes: sys-tematic review. BMJ. 2011; 343:d7163. doi: 10.1136/bmj.d7163 PMID: 22123912

17. Hu FB, Satija A, Manson JE. Curbing the diabetes pandemic. The need for global policy solutions.JAMA. 2015: 313:2319–2320. doi: 10.1001/jama.2015.5287 PMID: 25996138

18. Ardisson Korat AV, Willett WC, Hu FB. Diet, Lifestyle, and Genetic Risk Factors for Type 2 Diabetes: AReview from the Nurses’ Health Study, Nurses’ Health Study 2, and Health Professionals’ Follow-UpStudy. Curr Nutr Rep. 2014; 3:345−354. PMID: 25599007

19. O'Connor LM, Lentjes MA, Luben RN, Khaw KT, Wareham NJ, Forouhi NG. Dietary dairy productintake and incident type 2 diabetes: a prospective study using dietary data from a 7-day food diary. Dia-betologia. 2014; 57:909−917. doi: 10.1007/s00125-014-3176-1 PMID: 24510203

20. Bhupathiraju SN, Pan A, Manson JE, Willett WC, van DamRM, Hu FB. Changes in coffee intake andsubsequent risk of type 2 diabetes: three large cohorts of US men and women. Diabetologia. 2014;57:1346−1354. doi: 10.1007/s00125-014-3235-7 PMID: 24771089

21. Martinez-Gonzalez MA, Sanchez-Villegas A, De Irala J, Marti A, Martinez JA. Mediterranean diet andstroke: objectives and design of the SUN project. Seguimiento Universidad de Navarra. Nutr Neurosci.2002; 5:65−73. PMID: 11929200

22. Segui-Gomez M, de la Fuente C, Vazquez Z, de Irala J, Martinez-Gonzalez MA. Cohort profile: the'Seguimiento Universidad de Navarra' (SUN) study. Int J Epidemiol. 2006; 35:1417−1422. PMID:17060332

23. Willett W. Nutritional epidemiology. 2nd ed. New York: Oxford University Press, 1998.

24. Martin-Moreno JM, Boyle P, Gorgojo L, Maisonneuve P, Fernandez-Rodriguez JC, Salvini S, et al.Development and validation of a food frequency questionnaire in Spain. Int J Epidemiol. 1993; 22:512−519. PMID: 8359969

25. Fernandez-Ballart JD, Pinol JL, Zazpe I, Corella D, Carrasco P, Toledo E, et al. Martínez-GonzálezMA, Salas-Salvadó J, Martín-Moreno JM. Relative validity of a semi-quantitative food-frequency ques-tionnaire in an elderly Mediterranean population of Spain. Br J Nutr. 2010; 103:1808−1816. doi: 10.1017/S0007114509993837 PMID: 20102675

26. de la Fuente-Arrillaga C, VazquezRuiz Z, Bes-Rastrollo M, Sampson L, Martinez-Gonzalez MA. Repro-ducibility of an FFQ validated in Spain. Public Health Nutr. 2010; 13:1364−1372. doi: 10.1017/S1368980009993065 PMID: 20105389

27. Moreiras O, Carbajal A, Cabrera L, Cuadrado C Tablas de composición de alimentos [Food composi-tion tables]. 2013; 16th ed. Madrid: Piramide.





http://dx.doi.org/10.1016/S2213-8587(14)70057-9



http://dx.doi.org/10.1016/S0140-6736(09)61457-4

http://dx.doi.org/10.1016/S0140-6736(09)61457-4


http://dx.doi.org/10.1136/bmj.d7163





http://dx.doi.org/10.1007/s00125-014-3176-1


http://dx.doi.org/10.1007/s00125-014-3235-7





http://dx.doi.org/10.1017/S0007114509993837

http://dx.doi.org/10.1017/S0007114509993837


http://dx.doi.org/10.1017/S1368980009993065

http://dx.doi.org/10.1017/S1368980009993065


28. Mataix Verdu J. Tabla de composición de alimentos [Food composition tables]. 2009; 5th ed. Granada:Universidad de Granada.

29. Hu FB, Stampfer MJ, Rimm E, Ascherio A, Rosner BA, Spiegelman D, et al. Dietary fat and coronaryheart disease: a comparison of approaches for adjusting for total energy intake and modeling repeateddietary measurements. Am J Epidemiol. 1999; 149:531−540. PMID: 10084242

30. Greenland S. Analysis of polytomous exposures and outcomes. In Modern epidemiology. 3rd ed. Roth-man KJ, Greenland S, Lash TL, Eds. Philadelphia, Lippincott Williams andWilkins. 2008, p. 303−304.

31. Dominguez LJ, Bes-Rastrollo M, de la Fuente-Arrillaga C, Toledo E, Beunza JJ, Barbagallo M, et al.Similar prediction of decreased total mortality, diabetes incidence or cardiovascular events using rela-tive- and absolute-component Mediterranean diet score: the SUN cohort. Nutr Metab Cardiovasc Dis.2013; 23:451−458. doi: 10.1016/j.numecd.2011.10.009 PMID: 22402062

32. Schröder H, Fitó M, Estruch R, Martínez-González MA, Corella D, Salas-Salvadó J, et al. A shortscreener is valid for assessing Mediterranean diet adherence among older Spanish men and women. JNutr. 2011; 141:1140−1145. doi: 10.3945/jn.110.135566 PMID: 21508208

33. Martinez-Gonzalez MA, de la Fuente-Arrillaga C, Nunez-Cordoba JM, Basterra-Gortari FJ, Beunza JJ,Vazquez Z, et al. Adherence to Mediterranean diet and risk of developing diabetes: prospective cohortstudy. BMJ. 2008; 336:1348−1351. doi: 10.1136/bmj.39561.501007.BE PMID: 18511765

34. Bes-Rastrollo M, Perez Valdivieso JR, Sanchez-Villegas A, Alonso A, Martinez-Gonzalez MA. Valida-ción del peso e índice de masa corporal auto-declarados de los participantes de una cohorte de gra-duados universitarios. [Validation of self-reported weight and body mass index of participants in acohort of university graduates] Rev Esp Obes. 2005; 3:183−189.

35. Martinez-Gonzalez MA, Lopez-Fontana C, Varo JJ, Sanchez-Villegas A, Martinez JA. Validation of theSpanish version of the physical activity questionnaire used in the Nurses' Health Study and the HealthProfessionals' Follow-up Study. Public Health Nutr. 2005; 8:920−927. PMID: 16277809

36. Ainsworth BE, Haskell WL, Whitt MC, Irwin ML, Swartz AM, Strath SJ, et al. Compendium of physicalactivities: an update of activity codes and MET intensities. Med Sci Sports Exerc. 2000; 32:S498−S504. PMID: 10993420

37. Trichopoulou A, Costacou T, Bamia C, Trichopoulos D. Adherence to a Mediterranean diet and survivalin a Greek population. N Engl J Med. 2003; 348:2599−2608. PMID: 12826634

38. InterAct Consortium. Adherence to predefined dietary patterns and incident type 2 diabetes in Euro-pean populations: EPIC-InterAct Study. Diabetologia. 2014; 57:321−333. doi: 10.1007/s00125-013-3092-9 PMID: 24196190

39. Whincup PH, Donin AS. Cereal fibre and type 2 diabetes: time now for randomised controlled trials?Diabetologia. 2015; May 27 [Epub ahead of print].

40. Nowak C, Ingelsson E, Fall T. Use of type 2 diabetes risk scores in clinical practice: a call for action.Lancet Diabetes Endocrinol. 2015; 3:166−167. doi: 10.1016/S2213-8587(14)70261-X PMID:25636405

41. Phillips CM, Kearney PM, McCarthy VJ, Harrington JM, Fitzgerald AP, Perry IJ. Comparison of diabe-tes risk score estimates and cardiometabolic risk profiles in a middle-aged Irish population. PLoS One.2013; 8:e78950. doi: 10.1371/journal.pone.0078950 PMID: 24236074

42. Schmid R, Vollenweider P, Waeber G, Marques-Vidal P. Estimating the risk of developing type 2 diabe-tes: a comparison of several risk scores: the Cohorte Lausannoise study. Diabetes Care. 2011;34:1863−1868. doi: 10.2337/dc11-0206 PMID: 21636799

43. Imamura F, Micha R, Khatibzadeh S, Fahimi S, Shi P, Powles J, et al. on behalf of the Global Burden ofDiseases Nutrition and Chronic Diseases Expert Group (NutriCoDE). Dietary quality among men andwomen in 187 countries in 1990 and 2010: a systematic assessment. Lancet Glob Health. 2015; 3:e132−142. doi: 10.1016/S2214-109X(14)70381-X PMID: 25701991

44. Ramachandran A, Snehalatha C, Mary S, Mukesh B, Bhaskar AD, Vijay V. The Indian Diabetes Pre-vention Programme shows that lifestyle modification and metformin prevent type 2 diabetes in AsianIndian subjects with impaired glucose tolerance (IDPP-1). Diabetologia. 2006; 49:289−297. PMID:16391903

45. Colditz GA, Willett WC, Rotnitzy A, Manson JE. Weight gain as a risk factor for clinical diabetes mellitusin women. Ann Intern Med. 1995; 122:481−486. PMID: 7872581

46. Must A, Spadano J, Coakley EH, Field AE, Colditz G, Dietz WH. The disease burden associated withoverweight and obesity. JAMA. 1999; 282:1523−1529. PMID: 10546691

47. Sinclair A, Dunning T, Rodriguez-Mañas L. Diabetes in older people: new insights and remaining chal-lenges. Lancet Diabetes Endocrinol. 2015; 3:275−285. doi: 10.1016/S2213-8587(14)70176-7 PMID:25466523




http://dx.doi.org/10.1016/j.numecd.2011.10.009


http://dx.doi.org/10.3945/jn.110.135566


http://dx.doi.org/10.1136/bmj.39561.501007.BE





http://dx.doi.org/10.1007/s00125-013-3092-9

http://dx.doi.org/10.1007/s00125-013-3092-9


http://dx.doi.org/10.1016/S2213-8587(14)70261-X


http://dx.doi.org/10.1371/journal.pone.0078950


http://dx.doi.org/10.2337/dc11-0206


http://dx.doi.org/10.1016/S2214-109X(14)70381-X





http://dx.doi.org/10.1016/S2213-8587(14)70176-7


48. Morley JE, Vellas B, van Kan GA, Anker SD, Bauer JM, Bernabei R, et al. Frailty consensus: a call toaction. J AmMed Dir Assoc. 2013; 14:392−397. doi: 10.1016/j.jamda.2013.03.022 PMID: 23764209

49. Tobias DK, Pan A, Jackson CL, O’Reilly EJ, Ding EL, Willett WC, et al. Body-mass index and mortalityamong adults with incident type 2 diabetes. N Engl J Med. 2014; 370:233−244. doi: 10.1056/NEJMoa1304501 PMID: 24428469



http://dx.doi.org/10.1016/j.jamda.2013.03.022


http://dx.doi.org/10.1056/NEJMoa1304501

http://dx.doi.org/10.1056/NEJMoa1304501


researcharticle associationofadietaryscorewithincident...

Documents